The Role of Traffic Volume in Habitat Connectivity and Mortality Sandra Jacobson June 2007.

The Role of Traffic Volume in The Role of Traffic Volume in Habitat Connectivity and Habitat Connectivity and

MortalityMortality

Sandra JacobsonSandra Jacobson

June 2007June 2007

Most animal species make Most animal species make movements of at least 4 typesmovements of at least 4 types

Home range (regular movements within a fixed Home range (regular movements within a fixed area)area)Migratory (seasonal movements between Migratory (seasonal movements between breeding and non-breeding habitats)breeding and non-breeding habitats)Dispersal (movements by juveniles away from Dispersal (movements by juveniles away from the natal area)the natal area)Geographic range shifts (long-term movements Geographic range shifts (long-term movements in response to climatic or environmental in response to climatic or environmental changes)changes)From Hunter 1997From Hunter 1997

Natal dispersal movementsNatal dispersal movements

Increasing evidence that this movement type is Increasing evidence that this movement type is critical to maintain genetic and demographic critical to maintain genetic and demographic exchange among established populations (see exchange among established populations (see de Maynardier and Hunter 2000 p 62)de Maynardier and Hunter 2000 p 62)

Roads of all types filter demographic groups Roads of all types filter demographic groups differentlydifferently

Dispersing and migrating small mammals cross Dispersing and migrating small mammals cross roads more than resident ones (Swihart and roads more than resident ones (Swihart and Slade 1984) Slade 1984)

Traffic Volume as a Predictor of Traffic Volume as a Predictor of ImpactsImpacts

Volume is collected consistently with Volume is collected consistently with national standardsnational standards

DOTs understand itDOTs understand it

Good conceptual as well as empirical Good conceptual as well as empirical basisbasis– More research needed to verify species-More research needed to verify species-

specific effectsspecific effects

Definition of Traffic VolumeDefinition of Traffic Volume

Volume=Speed x DensityVolume=Speed x Density– Volume=vehicles/timeVolume=vehicles/time– Speed=distance/timeSpeed=distance/time– Density=vehicles/distanceDensity=vehicles/distance

VOLUME (Vehicles/time)=VOLUME (Vehicles/time)=Speed (distance/time) x Density (vehicles/distance)Speed (distance/time) x Density (vehicles/distance)

At LOW traffic intensity the small proportion of fauna casualties and animals At LOW traffic intensity the small proportion of fauna casualties and animals repelled causes limited impact on the proportion of animals successfully crossing a repelled causes limited impact on the proportion of animals successfully crossing a barrier. barrier. At MEDIUM traffic intensity casualties are high, the number of animals repelled by At MEDIUM traffic intensity casualties are high, the number of animals repelled by the infrastructure increases and the proportion of successful crossings decreases. the infrastructure increases and the proportion of successful crossings decreases. At HIGH traffic intensity a large proportion of animals are repelled and despite a At HIGH traffic intensity a large proportion of animals are repelled and despite a lower proportion of fauna casualties there is only a small proportion of successful lower proportion of fauna casualties there is only a small proportion of successful crossings. (Graph redrawn from Andreas Seiler, unpublished, text from COST-341 crossings. (Graph redrawn from Andreas Seiler, unpublished, text from COST-341 Handbook)Handbook)

0

10

20

30

40

50

60

70

80

90

100

Mean Traffic

Per

cent

age

Killed

Successful

Repelled

Two Major Types of Response to Two Major Types of Response to Traffic VolumeTraffic Volume

Large swift animalsLarge swift animals– usually smart and exhibit avoidance behaviorusually smart and exhibit avoidance behavior– >>Barrier effect predominates>>Barrier effect predominates

Small slow animalsSmall slow animals– often not smart and move regardless of often not smart and move regardless of

intimidation intimidation – Easy to mathematically predict complete Easy to mathematically predict complete

barrierbarrier– >>Mortality effect predominates>>Mortality effect predominates

Implications for Interpretation of Implications for Interpretation of Research ResultsResearch Results

Impacts do not increase linearlyImpacts do not increase linearly

Traffic volumes must be identified by Traffic volumes must be identified by appropriate measuresappropriate measures– Vehicles/hr for the activity period of the Vehicles/hr for the activity period of the

research speciesresearch species

Past research has been sloppy on terms Past research has been sloppy on terms and categories of volumeand categories of volume– Volume measurements need to be accurate Volume measurements need to be accurate

and precise for road in questionand precise for road in question– Define ‘high’ volumeDefine ‘high’ volume

Movement GroupsMovement Groups

Mathematically, traffic volume is Mathematically, traffic volume is straightforward:straightforward:– Speed of animals vs gap in trafficSpeed of animals vs gap in traffic

Movement groups allow for an organizing Movement groups allow for an organizing principleprinciple– Can be used across regions Can be used across regions

Research on the Relationship of Research on the Relationship of Traffic Volume to ImpactsTraffic Volume to Impacts

Van Langevelde and JaarsmaVan Langevelde and Jaarsma

Hels and BuchwaldHels and Buchwald

ArescoAresco

WallerWaller

Dodd and GagnonDodd and Gagnon

Probability of Successful Crossing: Probability of Successful Crossing: Parameters (Van Langevelde and Parameters (Van Langevelde and

Jaarsma 2004)Jaarsma 2004)Traffic volumeTraffic volumeTraversing speed of animalsTraversing speed of animalsBody length (size)Body length (size)Road widthRoad widthLarger spp that travel rapidly are the least vulnerable to roadkill Larger spp that travel rapidly are the least vulnerable to roadkill per crossing attempt (compared to other combinations)per crossing attempt (compared to other combinations)Because usually there are fewer large animals than smaller Because usually there are fewer large animals than smaller ones, large animal populations may be affected by absolute ones, large animal populations may be affected by absolute number losses compared to smaller animalsnumber losses compared to smaller animalsRoad crossings during daily movements as compared with Road crossings during daily movements as compared with fewer dispersal movements increase prob of mortality because fewer dispersal movements increase prob of mortality because of increased exposure to riskof increased exposure to risk

Probability of successful crossing Probability of successful crossing (van Langevelde and Jaarsma 04)(van Langevelde and Jaarsma 04)Traffic volume and traversing speed have largest Traffic volume and traversing speed have largest effect on probability of successeffect on probability of successIncrease in volume dramatically reduces P for Increase in volume dramatically reduces P for slow animals regardless of body sizeslow animals regardless of body sizeModel can be used to compare changes in Model can be used to compare changes in conditions conditions – Effectiveness monitoring of mitigation measuresEffectiveness monitoring of mitigation measures– Predicted future impacts from changed conditions Predicted future impacts from changed conditions

such as projected volume increasessuch as projected volume increases– See van Langevelde and Jaarsma 1997 See van Langevelde and Jaarsma 1997

(proceedings) for example(proceedings) for example

Determination of Probability of Determination of Probability of successful road crossingsuccessful road crossing

Sensitivity analysis: Jaarsma 04Sensitivity analysis: Jaarsma 04

Exact determination of pavement width, Exact determination of pavement width, traversing speed of animal, traffic volume traversing speed of animal, traffic volume is needed when:is needed when:– animals are slow or…animals are slow or…– traffic volume is hightraffic volume is high

Exact determination of traversing speed Exact determination of traversing speed and body length is NOT necessary for and body length is NOT necessary for larger mammals for roads with lower larger mammals for roads with lower volume or pavement widthvolume or pavement width

Number of Roadkills Number of Roadkills Van Langevelde and Jaarsma 2004Van Langevelde and Jaarsma 2004

Number of roadkills in a given time period Number of roadkills in a given time period estimated by estimated by

D=(1-P)KD=(1-P)K

D= number of roadkillsD= number of roadkills

P= probability of successfully crossing P= probability of successfully crossing onceonce

K=number of crossing attempts K=number of crossing attempts

Hels and Buchwald 2001: Hels and Buchwald 2001: Probability of small animals Probability of small animals

crossing road alivecrossing road alivep = e-Na/v

Where p is the probability of a successful crossing of one individual for one crossing attempt; N is the traffic volume per unit time (in AADT, Average Annual Daily Traffic); a is the combined kill zone of the animal size and vehicle surface; and v is the velocity of the animal (Hels and Buchwald 2001). This simplified equation assumes a perpendicular crossing thus it overestimates successful crossings. Jaeger and Fahrig (2004) determined that when the probability of successful crossing fell below 0.2, population persistence increased in all cases where access to a highway was restricted; therefore this value seems reasonable to define a complete barrier.

Differences between H&B and Differences between H&B and VL&J Probability ApproachesVL&J Probability Approaches

VL&J VL&J – Uses traffic flow theoryUses traffic flow theory– consider entire paved road width to be effective killing consider entire paved road width to be effective killing

surface (approximately true for large animals)surface (approximately true for large animals)– assumes traffic operates on a Poisson distribution assumes traffic operates on a Poisson distribution

which limits range of use to ca. 5000 AADTwhich limits range of use to ca. 5000 AADT– Does not consider crossing angleDoes not consider crossing angle

H&B H&B – Consider killing width based on tire size and animal Consider killing width based on tire size and animal

size because some animals can be on the road size because some animals can be on the road surface with a passing vehicle and survivesurface with a passing vehicle and survive

– Range of applicable AADT’s not stated but model is Range of applicable AADT’s not stated but model is limited to same traffic flow theory rangelimited to same traffic flow theory range

– Can account for crossing angle where knownCan account for crossing angle where known

Waller et al 2005 Probability of Waller et al 2005 Probability of successfully crossing roadsuccessfully crossing road

Also uses traffic flow theory Also uses traffic flow theory

Couched in more difficult to use format but Couched in more difficult to use format but the idea is the same: the longer an animal the idea is the same: the longer an animal is on the road, the more risk it incursis on the road, the more risk it incurs

Dodd and Gagnon in ArizonaDodd and Gagnon in Arizona

Ongoing researchOngoing research

Radiotelemetered elkRadiotelemetered elk

Highest collision rate is on weekdays; Highest collision rate is on weekdays; highest traffic is on weekendshighest traffic is on weekends

Difference in traffic volume is only about Difference in traffic volume is only about 3000 vehicles per day (i.e. 5000-8000)3000 vehicles per day (i.e. 5000-8000)

Animals travel farther to cross at high Animals travel farther to cross at high traffic volume timestraffic volume times

Traffic Calming (1)Traffic Calming (1)Definition: adaptation of roads to reduce vehicle Definition: adaptation of roads to reduce vehicle conflicts with other road usersconflicts with other road users

Traffic calming can enhance permeability by Traffic calming can enhance permeability by rearranging traffic so that volume on important rearranging traffic so that volume on important roads is reducedroads is reduced

Concept is important because volume is biggest Concept is important because volume is biggest impact, so volume can be managed for more impact, so volume can be managed for more favorable conditions for wildlifefavorable conditions for wildlife

Can rearrange traffic so that people can access Can rearrange traffic so that people can access needed areas but all roads are not increased to needed areas but all roads are not increased to high levels of servicehigh levels of service

Traffic Calming (2)Traffic Calming (2)‘‘Rat Running’ is through traffic using lower level Rat Running’ is through traffic using lower level of service roads to avoid higher traffic areas and of service roads to avoid higher traffic areas and to save time or distance (Jaarsma 97)to save time or distance (Jaarsma 97)Avoid ‘rat running’ by education and maintaining Avoid ‘rat running’ by education and maintaining level of service at lower levellevel of service at lower levelSee Jaarsma in CantersSee Jaarsma in CantersSee Jaarsma and Willems 2002See Jaarsma and Willems 2002

***This is a really good reason to ***This is a really good reason to mitigate very low volume roads, mitigate very low volume roads, because then when traffic is on low because then when traffic is on low volume roads they are essentially volume roads they are essentially non-barriers or mortality risksnon-barriers or mortality risks

Ausbau vor NeubauAusbau vor Neubau

Building up existing roads (Ausbau) is less Building up existing roads (Ausbau) is less impactive to wildlife than building new impactive to wildlife than building new roads (Neubau)roads (Neubau)

See Jaeger and Fahrig (2004) See Jaeger and Fahrig (2004)

Selected Online ResourcesSelected Online Resources

Interactive online model for road effects Interactive online model for road effects http://www.nls.ethz.ch/roadmodel/index.htmhttp://www.nls.ethz.ch/roadmodel/index.htm

Least cost pathway algorithm example Least cost pathway algorithm example http://www.grizzlybear.org/leastcostpath.htmhttp://www.grizzlybear.org/leastcostpath.htm

http://www.biodiversitypartners.org/http://www.biodiversitypartners.org/ – Conservation design section: Conservation design section:

http://www.biodiversitypartners.org/habconser/cnd/inhttp://www.biodiversitypartners.org/habconser/cnd/index.shtmldex.shtml

– Graphics for conservation design principles: Graphics for conservation design principles: http://www.biodiversitypartners.org/habconser/cnd/prihttp://www.biodiversitypartners.org/habconser/cnd/principles.shtmlnciples.shtml